Resistance coefficient of tee with capped end 1

[ARenko]

We have a "tee" (it's an API 6A block - see ling below) that we use as an elbow, by putting a flange on one outlet of the tee. Modeling this as a 90 degree miter bend results in significant reduction in the capacity of our control valve and I'm wondering if it's really appropriate to model it that way, or maybe it's closer to a LR or SR 90. Anyone know of any resistance coefficient for tee with one end capped?

Here's an API cross...

https://www.woodcousa.com/crs-tee/5185x5185t.htm

Imagine if the top outlet is not there, the flow comes in from the left, the right outlet is flanged, and the flow exits the bottom.

 

[1503-44]

Well that's somenting that Crane never thought of. Me either. And nobody has ever asked that question anywhere that I know of.

I think you should treat it like a miter, at least until somebody runs a cfd on it. An ell would be too smooth, IMO.

A black swan to a turkey is a white swan to the butcher ... and to Boeing.

 

[LittleInch]

These are a bit odd. Normally when I've seen a "target tee", it's been a standard tee with a flanged end.

This cross is rather more brutal than a swept tee and the open hole will have an impact on the flow so even a SR elbow I think is optimistic.

Is there no other way? Why not use the choke valve to turn the flow 90 degrees?

I don't think I've ever seen a 5000psi cross in service before tbh in this mode.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[GBTorpenhow]

Crane TP410 (the more recent versions) gives a method for estimating resistance factors of converging and diverging tees with various branch/run flow ratios.

For a straight tee with flow into the branch and all coming out one run, Crane gives K = 1.1 and with flow into one run and all coming out of the branch, K = 1.7.

 

[1503-44]

GBT is right. The capped end stumped me, but then I thought why not just consider no flow into that one. Use it.

A black swan to a turkey is a white swan to the butcher ... and to Boeing.

 

[ARenko]

LittleInch,
The tee turns the flow back in the direction it came from. Flow comes into the choke from below, turns 90 deg right (choke is 90 deg valve), then the tee turns the flow 90 deg down. There is pipe in between, but it's distance is set by a gate valve on a bypass line. For two pipe sizes the tee entry flange is 1" and 2" inside 6 pipe diameters from choke exit flange, but for one size it's well into it.

GBTorpenhow,
Thanks, I missed that section. In this case the flow is into the run and all flow out the branch. A quick look at the graphical representation suggests K=.68, but I haven't read through the section yet. Am I missing something? None of the graphs go as high as 1.7.

 

[LittleInch]

It's very surprising that you need API 5000 piping downstream of your choke.

Normally the choke gets the pressure down to something like ASME class 600 or 900 with the wing valves acting as your ESD valves.

Flow into a choke is normally in from the side and out the bottom.

It all sounds a bit odd to me but I can't see what you're looking at...

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[ARenko]

I agree, 5000 psi rating is not needed downstream of the choke. Heck, our system really doesn't need to be 5K in the first place, as upstream pressure limitation is 2K, but this is the way it is.

The application is a choke manifold in managed pressure drilling. I've suggested we have a spec break d/s of the choke, but customers insist (or at least that's what they've been sold on). Downstream we have a meter skid with all 5K components except the meter that is 600# - 1500#. And then immediately after that we're down to 300# or 600# piping from the rig to the rest of their process. Drilling guys don't think the same as process guys. I think the big concern is plugging, even though we have relief valves upstream.

There's no normal in drilling choke orientation - if there was I think it would be in/ out on the same horizontal plane. We go vertical to minimize footprint.

That said I'm not sure what it has to do with my question, but I appreciate your interest to understand better.

ETA: we do sometimes now put the meter upstream to improve reading when gas is present - in that case it's the 1500# meter. Still don't need 5K based on upstream limitations though.

 

[GBTorpenhow]

ARenko - I don't have the modern TP410 in front of me currently, at one point I built all the formulas given in that section into a spreadsheet. Run through it with Beta = 1 and Q_branch/Q_combined = 1 and see what you get.

 

[LittleInch]

Well it seemed you were creating a problem where one shouldn't exist. So sometimes it's more interesting to look behind the immediate issue to suggest other ways to solve the problem

"but this is the way it is" is an abdication of design IMHO. It makes zero sense to have equipment like 6A 5K crosses when another part of the same system is within an ASME class limit.

Then just use a standard LR elbow and everything is better and costs less.

Drillers just need to be told to go drill sometimes and leave the design to others.

Must chokes I've ever seen are 90 degrees, for me usually in at the side out the base but can be the other way around.

But your question is answered I think.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[georgeverghese]

"Flow comes into the choke from below, turns 90 deg right (choke is 90 deg valve), then the tee turns the flow 90 deg down."

Standard Company practice at an oil/gas supermajor : Never orient a choke valve in high erosion service with flow going into choke from the bottom. Always make the flow go in through the side where flow momentum and sand erosion will not impact on the valve stem and packing. So suggest moving the choke to the current downstream tee.

 

[ARenko]

It's not uncommon in this type of equipment to orient the choke with flow coming in from bottom. These are drilling chokes that are much more robust than typical production chokes - they process drilling mud with cuttings and sometimes small amounts of gas. Cavitation is unavoidable due to the large range of operating conditions of the drilling process. We have 70+ systems (most with one horizontal and one vertical choke) and haven't noticed any significant difference between the two orientations. But as our line size and number of choke legs grows to meet higher flow rate requirements newer systems are all vertical in order to keep the size manageable for transport and finding a spot for it on the rig.

Anyway, thanks to everyone for all your comments.

 

[SPDL310]

LittleInch,

I have not designed an MPD manifold myself, but have put together 3rd party review packages for multiple systems that have been put together. Because theses systems are sometimes built as modular rental equipment they are certified individually. I have seen them most often split into buffer, choke and meter manifold packages. The choke and buffer packages are normally rated 5K for versatility, though I have seen some built for 2K service. ABS and DNV like to see a consistent pressure rating across the entire skid with spec breaks at the connection between skids. The meter manifolds are usually 2K components with pressure rating de-rated to #600 for the meter, but less commonly a #1500 meter is de-rated to 2K for the piping.

As to the original question of the thread, I believe that the resistance should fall out somewhere between between a mitered 90 and a SR elbow. Probably closer to the mitered bend than the SR. The cushion block tees in 16C behave similarly. You will have a majority of the flow that make the 90degree turn, but another portion will enter the cushion and make a 270 degree turn.

 

[ARenko]

LittleInch,

"Well it seemed you were creating a problem where one shouldn't exist." It is not really a problem, I just wanted to make sure my calculations were accurate.

And we have convinced some customers to go with lower pressure ratings, but if they want to pay for higher pressure ratings than they need and they pay for it, so be it.